报告题目：Feature Extraction and Dimension Reduction of Color Images in Quaternion Domain
A quaternion number is a number system that contains a real part and three imagery parts. Putting three color channels of a color image into three imagery parts, a color image pixel can be encoded as a pure quaternion number. The quaternion representation of a color image maintains the spatial correlations of color image pixels while encoding the cross-channel relationship of the color image. In this talk. I will focus on color image processing in quaternion domain. After a brief introduction of quaternion algebra, I will present several feature extraction and dimension reduction methods for color image classification and recognition.
Dr. Yicong ZHOU received his B.S. degree from Hunan University, Changsha, China, and his M.S. and Ph.D. degrees from Tufts University, USA, all in electrical engineering. Dr. Zhou is currently an Associate Professor and Director of the Vision and Image Processing Laboratory in the Department of Computer and Information Science at University of Macau, Macau, China. His research interests include chaotic systems, multimedia security, image processing and understanding, and machine learning. Dr. Zhou serves as Associate Editors for Signal Processing: Image Communication, Neurocomputing and the Journal of Visual Communication and Image Representation, and a co-chair of IEEE SMC Technical Committee on Cognitive Computing. Dr. Zhou is a senior member of IEEE and was a recipient of the third price of Macau Natural Science Award in 2014.